diff options
author | jkim <jkim@FreeBSD.org> | 2016-05-25 23:06:52 +0000 |
---|---|---|
committer | jkim <jkim@FreeBSD.org> | 2016-05-25 23:06:52 +0000 |
commit | 45ae491494493d7935a06c1ad0e655d974d1cf22 (patch) | |
tree | 18b691be837ea394d57aa5fb5d89808be9d3964c /sys/amd64/amd64/pmap.c | |
parent | dad0f4b1ee3df3366262c758edb7d3b2b37678f4 (diff) | |
download | FreeBSD-src-45ae491494493d7935a06c1ad0e655d974d1cf22.zip FreeBSD-src-45ae491494493d7935a06c1ad0e655d974d1cf22.tar.gz |
Both Clang and GCC cannot generate efficient reserve_pv_entries().
http://docs.freebsd.org/cgi/mid.cgi?552BFEB2.8040407
Re-implement it entirely in inline assembly not to let compilers do silly
spilling to memory. For non-POPCNT case, use newly added bit_count(3).
Reported by: alc
Reviewed by: alc, kib
Differential Revision: https://reviews.freebsd.org/D6541
Diffstat (limited to 'sys/amd64/amd64/pmap.c')
-rw-r--r-- | sys/amd64/amd64/pmap.c | 31 |
1 files changed, 15 insertions, 16 deletions
diff --git a/sys/amd64/amd64/pmap.c b/sys/amd64/amd64/pmap.c index 7132c57..cc76974 100644 --- a/sys/amd64/amd64/pmap.c +++ b/sys/amd64/amd64/pmap.c @@ -104,6 +104,7 @@ __FBSDID("$FreeBSD$"); #include "opt_vm.h" #include <sys/param.h> +#include <sys/bitstring.h> #include <sys/bus.h> #include <sys/systm.h> #include <sys/kernel.h> @@ -585,7 +586,7 @@ static caddr_t crashdumpmap; static void free_pv_chunk(struct pv_chunk *pc); static void free_pv_entry(pmap_t pmap, pv_entry_t pv); static pv_entry_t get_pv_entry(pmap_t pmap, struct rwlock **lockp); -static int popcnt_pc_map_elem_pq(uint64_t elem); +static int popcnt_pc_map_pq(uint64_t *map); static vm_page_t reclaim_pv_chunk(pmap_t locked_pmap, struct rwlock **lockp); static void reserve_pv_entries(pmap_t pmap, int needed, struct rwlock **lockp); @@ -3126,7 +3127,7 @@ retry: } /* - * Returns the number of one bits within the given PV chunk map element. + * Returns the number of one bits within the given PV chunk map. * * The erratas for Intel processors state that "POPCNT Instruction May * Take Longer to Execute Than Expected". It is believed that the @@ -3142,12 +3143,15 @@ retry: * 6th Gen Core: SKL029 */ static int -popcnt_pc_map_elem_pq(uint64_t elem) +popcnt_pc_map_pq(uint64_t *map) { - u_long result; + u_long result, tmp; - __asm __volatile("xorl %k0,%k0;popcntq %1,%0" - : "=&r" (result) : "rm" (elem)); + __asm __volatile("xorl %k0,%k0;popcntq %2,%0;" + "xorl %k1,%k1;popcntq %3,%1;addl %k1,%k0;" + "xorl %k1,%k1;popcntq %4,%1;addl %k1,%k0" + : "=&r" (result), "=&r" (tmp) + : "m" (map[0]), "m" (map[1]), "m" (map[2])); return (result); } @@ -3179,17 +3183,12 @@ retry: avail = 0; TAILQ_FOREACH(pc, &pmap->pm_pvchunk, pc_list) { #ifndef __POPCNT__ - if ((cpu_feature2 & CPUID2_POPCNT) == 0) { - free = bitcount64(pc->pc_map[0]); - free += bitcount64(pc->pc_map[1]); - free += bitcount64(pc->pc_map[2]); - } else + if ((cpu_feature2 & CPUID2_POPCNT) == 0) + bit_count((bitstr_t *)pc->pc_map, 0, + sizeof(pc->pc_map) * NBBY, &free); + else #endif - { - free = popcnt_pc_map_elem_pq(pc->pc_map[0]); - free += popcnt_pc_map_elem_pq(pc->pc_map[1]); - free += popcnt_pc_map_elem_pq(pc->pc_map[2]); - } + free = popcnt_pc_map_pq(pc->pc_map); if (free == 0) break; avail += free; |